Anthropic says Claude AI hacked three organisations during cyber tests

Anthropic says Claude AI hacked three organisations during cyber tests
It comes just days after rival OpenAI said rogue AI agents had breached other firms' networks.